HEATIT - Hashing EvAluation ToolkIT
Supported hashing methods:
AGH, BRE, CH, CPH, DSH, IsoH, ITQ, KLSH, LSH, SH, SpH, USPLH
You can get the following curves drawn by the Hamming ranking method:
- Recall - the number of retrieved samples curves
- Precision - the number of retrieved samples curves
- Precision - recall curves
- Download siftsmall.tar.gz, siftsmall.tar.gz and gist.tar.gz from here.
- Extract the data set you want to use to the corresponding folder.
- Run
heatit.mlapp
using MATLAB App Designer.
- Download siftsmall.tar.gz, siftsmall.tar.gz and gist.tar.gz from here.
- Extract the data set you want to use to the corresponding folder.
- Run
heatit.m
. - You can choose different
datasetCandi
,methodCandi
,numberOfPoint
andcodelength
parameters. Modifying any code ofheatit.m
does not affect the behavior ofheatit.mlapp
.
% Offline training
function [model, train_coded, train_time] = method_learn(trainset, codelength)
% Online search (encoding only)
function [test_coded, test_time] = method_compress(testset, model)
Heatit is licensed under the MIT License.